4 papers · 1 filter
SOM-VQ: Topology-Aware Tokenization for Interactive Generative Models
Alessandro Londei, Denise Lanzieri, Matteo Benati
Vector-quantized representations enable powerful discrete generative models but lack semantic structure in token space, limiting interpretable human control. We introduce SOM-VQ, a…
Inverting Self-Organizing Maps: A Unified Activation-Based Framework
Alessandro Londei, Matteo Benati, Denise Lanzieri +1
Self-Organizing Maps (SOMs) provide topology-preserving projections of high-dimensional data, yet their use as generative models remains largely unexplored. We show that the activa…
Lyapunov Learning at the Onset of Chaos
Matteo Benati, Alessandro Londei, Denise Lanzieri +1
Handling regime shifts and non-stationary time series in deep learning systems presents a significant challenge. In the case of online learning, when new information is introduced,…
Dreaming Learning
Alessandro Londei, Matteo Benati, Denise Lanzieri +1
Incorporating novelties into deep learning systems remains a challenging problem. Introducing new information to a machine learning system can interfere with previously stored data…